Digital Financial Reporting

Eiichi "Ed" Watanabe, an early evangelist for XBRL in Japan and globally, has set up a web site, XBRL House. This is how Watanabe-san describes his web site:

My intent of creating this website was to share what I have with respect to XBRL so that  the information included would be of some value for those who join the XBRL world. I thought it would be good for me to do that because I was so involved in XBRL and I am retired almost from XBRL, although I am still an advisor of XBRL Japan.

One thing I would point out on the site is the library.  I counted 20 books related to XBRL.

Stop by XBRL House and say hello to Ed!

I have mentioned Microsoft PowerPivot before. But, I did not have Excel 2010 at the time; now I do.  I did not know about the Windows Azure Marketplace back then.

PowerPivot is basically an Excel pivot table on steroids.  It is a free add in to Excel.  It is a beautiful looking interface, but it does take some time to figure out.

The Windows Azure Marketplace is, well, a market place where publishers of applications and data can sell there wares to people who want to buy those applications and data.  There is an Excel add in, also free, which lets you connect to these data sets.

This video what helpful in getting a connection to some free data which is available in the Azure Marketplace, basically letting you experiment with Azure and PowerPivot.

All this stuff is worth checking out. 

My question is: When will working with XBRL information be this easy?

The move to digital financial reporting will cause a number of very significant shifts.  One of these shifts relates to how metadata can be employed.  This shift is both an opportunity and a threat.  Most accountants and CPAs don't have a good enough grasp as to what metadata is or the role it plays.  Therefore, nor do they understand the side of the shift equation on which they will end up.

This is what I mean.

Let me use the external financial statement as an example.  Most external financial statements today are created using Microsoft Word.  I hear the number 85%.  I am not talking about the balance sheet, income statement, and maybe cash flow statement which might be generated from an accounting or ERP system.  I am talking about a complete financial statement.

So, how much does Microsoft Word know about financial statements?  You are probably thinking that this is a rather odd question, of course Word knows nothing about financial statements.  The person creating the financial statement is the one which knows about financial statements; they use their knowledge of financial reporting and US GAAP or IFRS to create a financial statement using Word.

That is exactly the problem.  In fact, it is two problems.  The first problem is that Word cannot help you create that financial statement and get it correct.  The second problem is that once the information is put into Word, because Word does not have any knowledge of the financial information within the financial statement; reusing that information involves humans, usually with domain knowledge, rekeying that information in order to make the information actionable.

The graphic below shows this has worked in the past (and likely how most people do this today):

But what if Word did understand financial statements?  Well, word my never understand external financial statements, but other applications will.  You could get Word to understand financial statements by using its macro language, which is VBA and actually quite powerful, but more likely other applications will be created.  And how might a software application understand a financial statement you might ask? Structured information and metadata. (To understand that statement is a journey, start here on that journey. This has the complete story. Once you see this in action, you will get it, get it.)

Basically information which you and other know about a domain such as financial reporting will be expressed in a form which a computer can understand.  That is what the XBRL medium does, it expresses information in a structured form so a computer can understand it.  Additional information, metadata, will be expressed which is helpful in working with that structured information.

Algorithms, or computer programs, will do stuff with that structured information and metadata.  Lots of stuff.  The graphic below shows this:

So two things will happen.  First, who can create information and how they create that information will change dramatically. Because the computer can help the user, a less skilled person can do the work because the "human knowledge" is now expressed in the metadata.  Second, automated reusing the information will become possible.

This will spiral, the possibilities widening and widening as more and more metadata and algorithms are created and employed.

What is the threat?  If your skill is memorizing and regurgitating information, this is a threat.  If your skill is rekeying information, this is a threat.  Basically, think of what numerically controlled (NC) machines did to the manufacturing process.  Robots build a lot of stuff today using algorithms and metadata which control the machines which churn out consistent, higher quality output than humans can generally create.

The opportunity? Creating algorithms, creating metadata, doing value-add analysis of all that structured, model-based information.

And these opportunities and threats are not limited to external financial reporting, or even financial reporting.  These same ideas can be applied to many, many other domains.

Stay tuned to my blog, it may help you understand and take advantages of the opportunities and not end up as road kill.

The best explanation of what actionable information is that I have found came from an article by John Alber, Delivering Actionable Information To Front-Line Lawyers:

If a friend tells you that you have something in your teeth, chances are you’ll visit a mirror and attend to the problem. That’s actionable information. It is information (1) from a trusted source, (2) about something that’s important to you, and (3) that, once known to you, will impel you to take action.

While the article talks about law firms, it has general aplicability in my view.

Ask yourself this question.  In your organization, how does the mass of information which you have available become actionable?  Is that process as effecient and as effective as it can be?  Chances are that the process involves lots of reports, spreadsheets, re-keying, etc.

Think about this. More later on how to improve this process.

The folks involved with EuroFiling have come up with an interesting idea called the "data points model" approach to modeling information in XBRL. As far as I know, this web page has a bunch of information relating to this approach.  I sat through a presentation which explained the data points model approach and this is what I understood.

Basically, there are three important pieces to this idea:

• Base item: The base item is the concept of that fact.
• Breakdown: The breakdown an [Axis] or dimension of that item.
• Data point: A data point is a reported fact or a "cell in a template".

So for example, base items might be something like: assets, liabilities, equity, revenues, expenses, gains, losses, transfers to owners, transfers from owners.

For the base item "assets", the breakdown might be cash and cash equivalents; receivables; property, plant and equipment, and so forth.

It appears that the goal of this approach is to achieve better flexibility in using the information and understanding facts which go together with other facts.

As part of this approach a "table linkbase" is created.  This piece I don't quite get yet.  What I do get is that no off the shelf software supports this table linkbase and it is not a global standard approach.

This data points model approach is worth watching.  I wish they would better explain this and provide prototypes so one can better evaluate the pros and cons of the approach.

#############################

Introduction to the Data Point Model (DPM)

Data Point Modeling

EIOPA Data Point Methodology

DPM as ISO Standard